Why Solar Plus Storage Hit 16% of US Solar Installations

In 2017, roughly one in fifty solar installations in the United States came with a battery attached. Today the figure is about one in six. According to a US Energy Information Administration analysis reported by Utility Dive, co-located solar and storage now makes up 16% of all solar installations, up from 2% in 2017.

That is an eightfold rise in under a decade, and it is easy to read it as simple enthusiasm for batteries. The more interesting reading is that developers are bundling the two because the grid, the tax code and the physics all reward it. This piece separates what the data say from what they might mean.

What the EIA figure actually says

Fact: The EIA reports that co-located solar and storage has grown from 2% of solar installations in 2017 to 16%. “Co-located” means a battery sits at the same site as a solar plant, usually behind a shared grid connection. The source is the EIA's own tracking of US generators, as summarized by Utility Dive.

Two caveats matter before anyone quotes the number. First, a share of “installations” can mean a share of projects or a share of megawatts, and those can differ a lot. A handful of very large Western projects can swing a capacity figure that a count of small sites would barely register. Check which one a given chart uses before comparing it to anything else.

Second, co-located does not mean co-operated. A battery on a solar site may charge from the grid, discharge on a schedule unrelated to the panels, or sit idle under a contract quirk. The label describes a shared address and often a shared meter, not a shared strategy.

Why developers bundle the two

Interpretation: Several incentives point the same way, and none of them requires anyone to love batteries.

The first is the interconnection queue. A solar project already holds, or is waiting for, a slot to connect to the grid. Adding a battery behind that same connection can be quicker than filing a new standalone request, because the project reuses existing wires and permits. As Greadly's look at the grid connection queue showed, access to the grid is often scarcer than the equipment itself.

The second is energy capture. Solar farms are often built with more panel capacity than inverter capacity, and the excess gets clipped on sunny middays. A battery on the DC side can soak up that clipped energy instead of discarding it. This is a standard engineering rationale, though how much it matters depends on each project's design.

The third is tax policy. Since the 2022 Inflation Reduction Act, standalone storage has been eligible for the federal investment tax credit, which removed one old reason to pair it with solar. Even so, summaries of the July 2025 budget law describe an earlier phase-out for wind and solar credits than for storage credits. A developer who adds storage may keep a stronger credit position for part of the project. That is an incentive worth checking against each project's actual filings, not a universal rule.

The midday problem that batteries solve

The deeper driver is what too much solar does to its own market. When many panels produce at once, midday power prices fall, and in some places they fall to zero or below. A project that can only sell at noon earns less as its neighbors multiply.

A battery lets the plant move some of that output to the evening, when demand peaks and prices are higher. This is the same logic behind why four-hour batteries now beat gas peakers on cost: shifting cheap daytime energy into expensive evening hours is the business. Co-location puts the shifting machine next to the source.

The alternative is wasted power. In India, for example, curtailment left solar output unused even while the country ran short, a problem covered in an earlier Greadly piece on India's wasted solar power. The conditions differ across countries, but the lesson travels: generation without a way to hold it eventually hits a ceiling.

What the 16% does not prove

Interpretation: The number shows that bundling is now normal for new projects. It does not show that the grid has enough storage, or that the storage is in the right places.

Co-located batteries sit where solar is cheap to build, which is not always where the grid is stressed. Regions facing evening shortfalls, such as the PJM territory discussed in an earlier analysis of PJM's power crunch, need capacity in specific zones. A standalone battery at a congested substation may help more than a co-located one in a sunny but well-supplied corner.

There is also a measurement gap. The headline share counts what has been built or reported, not how the batteries behave. Without operating data, a reader cannot tell whether these systems are cycling daily to cover the evening peak or mostly holding reserves. Public data on that behavior lag the construction data.

What to watch next

Prediction: These are judgments, not forecasts from the EIA. I expect the co-located share to keep rising while standalone storage also grows, because the two serve different needs. I would not expect a steady line. Tax-credit deadlines, tariffs on imported cells and the sourcing rules attached to federal credits can pull project timing forward or push it back.

Three indicators will say more than another share-of-installations chart:

• Capacity versus count: whether the 16% holds when measured in megawatts and megawatt-hours.
• Charging source: how often co-located batteries charge from their own panels rather than the grid.
• Dispatch hours: whether batteries discharge into the evening peak, or into whatever window a contract specifies.

The 2% to 16% climb is real, and it reflects a sensible response to crowded interconnection queues and cannibalized midday prices. But it is a count of how projects are being packaged. Whether that packaging keeps the lights on at 7 p.m. is a separate question, and the answer sits in operating data that have yet to catch up.